I recently updated the Tools section of the site with a refresh of the tools, companies, and technologies in finance today, and redid the screenshots and many of the blurbs on each company. One trend was clear: everyone was building AI into their value proposition, in terms of the primary value message, new features, core principles, or even shifts in strategy or product. The interest in AI is everywhere.
Back in December 2022 I wrote How will we use AI to build spreadsheet forecast models?:
It makes sense that incorporating AI into the process of building a spreadsheet forecast model can enable companies to benefit from the enhanced data analysis and predictive capabilities of AI. Gather financial and operational data - historical financial information such as income statements, balance sheets, and cash flow statements, as well as data on expected changes in the company's operations, such as new products or services, changes in pricing, or expansion into new markets - and train an AI model to use for forecasting and data analysis. Take public market data to create a tool to query for benchmarks and comparisons for types of businesses at specific stages. Train a model on your own company's data to create a tool to help forecast future sales, inventories, cash position.
Once we have an AI model, the results can then be incorporated into a spreadsheet forecast model, using cells as prompts to generate forward-looking projections or generate insights from historical data to inform the forecast assumptions and projections.
In addition to improving the accuracy and efficiency of the forecast model, AI could be used to automate certain aspects of the model-building process. Automatically update the model with new data as it becomes available. Automatically generate reports and presentations based on AI-powered inferences on the actual and forecasted data in the model. Dump data on a company into an AI model for it to figure out which statistical and forecasting methods to use, and why.
We're starting to see spreadsheet modeling introduce AI into their tools, notably Shortcut, a model building AI tool or agentic tool for Excel, and Excel's own Agent Mode. These tools extend the idea of "vibe coding" into "vibe working", the idea that you iterate on the model through back-and-forth prompts with the tool to adjust and build your model and analysis. While the benchmark performance of 50-60% for these tools don't sound great, it's actually not that far off human modeling performance, which is notoriously (or perhaps surprisingly) often less-than-accurate.
AI is coming into every finance tool, but what does that mean?
The shift from building to iterating
The introduction of AI agents into financial modeling represents a fundamental shift in how we work with forecasts. Instead of starting from scratch or copying templates, we're moving toward an iterative, conversational approach to model building. You describe what you need, the AI builds it, you review and refine through prompts, and the cycle continues until you have something that works.
This "vibe working" approach has real benefits. It can help less experienced modelers get started faster, reduce the time spent on repetitive formula writing, and enable faster experimentation with different assumptions and scenarios. For experienced modelers, it can accelerate the initial model setup, allowing more time for analysis and strategic thinking rather than mechanical spreadsheet construction.
The accuracy question
The 50-60% accuracy rate for these tools might seem concerning, but it's worth putting that in context. Human-built financial models are often riddled with errors, from simple formula mistakes to structural logic problems, and these problems often go unnoticed until they cause significant problems.
AI isn't perfect either. As I wrote in Don't trust your AI with your cap table, AI tools can generate models that sound authoritative but can still have fundamental subject knowledge and mathematical model errors. All errors matter: a misplaced decimal point, an incorrect reference, or a flawed assumption can cascade through a model and lead to poor decisions.
The key is to treat AI as a powerful assistant, not a replacement for human judgment. Use it to accelerate your work, but always verify the logic, check the formulas, and validate the outputs against your understanding of the business.
What this means for FP&A teams
For FP&A professionals, this shift means:
- Faster model creation: Less time building, more time analyzing
- Lower barriers to entry: Team members with less spreadsheet expertise can contribute more effectively
- More experimentation: Easier to test different scenarios and assumptions
- New skills required: Understanding how to prompt effectively and validate AI-generated work becomes critical
The tools are getting better, but they're not perfect. The best approach is to use AI to handle the mechanical aspects of model building while you focus on the strategic thinking, business understanding, and validation that ensures your forecasts are both accurate and useful.
The future of financial modeling isn't about AI replacing analysts, it's about AI augmenting their capabilities, allowing them to spend less time on formula writing and more time on the analysis and insights that drive better business decisions.
Questions about AI in forecasting, ask anytime.